ObjectivesIn order to clarify how different treeing lifetime and tree growth behavior were between epoxy resins and their nanocomposites, laboratory-prepared epoxy/boehmite alumina nanocomposites were subjected to local high electric filed using treeing test cells, and were evaluated on their treeing performances. Both results experimentally obtained are interpreted based on a multi-core model, and some contradiction in explanation is pointed out for further research. A crossover phenomenon was newly found and interpreted in terms of the multi-core model. ResultsTreeing phenomena in epoxy/alumina nanocomposite were experimentally investigated and explained to some extent in terms of a multi-core model. Role of nano-fillers was discussed and clarified for treeing resistance. Major results are as follows:(１) Alumina nano-fillers are effective to suppress both tree initiation and propagation. Optimum content of the nano-fillers is around 5 wt% at our current manufacturing technology.(２) It was found that there are two kinds of tree growth processes, i.e. thin channel formation and thick channel formation. At high field, the former is followed by the latter.( ３ ) It was elucidated that thin tree channels reach the opposite electrode at high field, but cause no breakdown in both pure epoxy and its nanocomposite. Treeing breakdown takes place, when thick channels bridge a dielectric specimen, as usually anticipated.(４) Tree channels are broadened from thin to thick by partial discharges (PD's) taking place inside the channels. PD products emerge onto the inner surface in nanocomposites, but no PD products are traced in pure epoxy.(５) A crossover phenomenon as shown in Fig. 1 is recognized in tree growth vs. applied voltage characteristics between pure epoxy and its nanocomposite. Selective paths for thin tree channel formation at high field are assumed in the nanocomposite for explanation.(６) Increase in tree initiation threshold is explained by the increase in the height of a barrier between the electrode and the dielectric. This increase is substantiated by the effect of charge redistribution that is considered to be formed near the metal electrode. This charge redistribution must be caused by the collective role of the electric double layers that each of the multi-cores in the multi-core model may possess.(７) Difference in the speed of thin channel formation at high electric field between pure epoxy and its nanocomposite is interpreted by the concept of rather straightforward selective paths. It is suggested that such paths will be effective with the repulsive force caused by charge tails of the electric double layers that the multi-cores of nano-fillers generate.(８) At medium electric field, nano-fillers or multi-cores are considered to work as high permittivity media to invite a tree tip filled with partially ionized conducting gas. This can explain longer tree growth time in nanocomposites than in pure epoxy.(９) Nanocomposites are more erosion-resistant inside the tree channels than pure epoxy, as clarified in s...

No abstract

No abstract

Solder bridging test to evaluate the compatibility for fine-pitch microsoldering Establishes a quantitative evaluation method on solder bridging in microsoldering using a printed wiring board with comb pattern conductors. Bridge tests were conducted by immersing the comb pattern board into a molten solder bath. The total length of solder bridge between conductors against the total length between conductors was measured as a measure of the occurrence of solder bridging. The occurrence of bridging depended on the number of immersions, flux activity including solid content, conductor spacing, solder bath temperature and solder composition. The increase in number of immersions enhanced bridging. The rosin flux without activators showed higher bridging than the activated flux. Sn-37Pb solder showed lower bridging than Sn-3.5Ag-5Bi solder. Solder bridging was found to be closely correlated with wettability; therefore, the improvement of wettability could be effective to suppress solder bridging. The proposed method is believed to be suitable for the quantitative evaluation of solder bridging.

No abstract